Project ReportΒΆ

Date: 2025-09-29

part1ΒΆ

Part 1ΒΆ

No description has been provided for this image

The Orthogonal images appear to be more flat against a background, while the perspective angles give more depth to the image. Specifically, I think more of the top part of the object gets shown. It is also interesting how much darker the perspective images are, although I suspect it is due to shadows. The final interesting piece is the perspective image seems to be all in line with eachother.

part2ΒΆ

Part 2ΒΆ

Question 1ΒΆ

Out[3]:
(0.0, 256.0)
No description has been provided for this image
Out[4]:
Text(0.5, 1.0, 'Equalized Image 3')
No description has been provided for this image

Question 2ΒΆ

No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image

In both of the pictures I took, the histogram equalization was relatively helpful. In the perspectove image in particular the equalization helped birghten the image and make the power button more clear. In the orthogonal image, equalization had a weird effect. It seems to just add more noise but it also texturized the image. I think it could be beneficial but maybe not necessarily worth the trade off. I think it was very interesting to see how the graphs looked post equalization, it backs up the idea that the process adds more noise, or variation between their heights.

No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image

part3ΒΆ

Part 3ΒΆ

Question 1ΒΆ

No description has been provided for this image
No description has been provided for this image

Question 2ΒΆ

No description has been provided for this image

It seems that Anisotropic diffusion is more adept at specifying pieces of an image, the Gaussian Smoothing is more of just a filter. Anisotropic Diffusion also does reduce the blurring of edges as Gaussian filters do. While also printing an exampolen that reduces this slightly to help identify smaller edges in the image.

Question 3ΒΆ

The easiest issue to identify is how important parameters are for Anisottopic diffusion, accidentally blurring important edges or not blurring noise enough are 2 comprehensible issues. However, it will also create staircasing artifacts. Not a huge deal but in many iterations can start to make the images look artificial.